Linear Regression Model was adopted where we can see 80% of data for training and 20% for testing by default so that model can be trained on unseen data

R squared value shows the variation in the data that our model explains - so our linear regression model explains 72.7391% of the variation in median value of occupied homes

Predictions yield a more tangible feel for how the model acts on the data, and we can extract mean squared error and root mean squared error scores which can potentially motivate F-Statistical Tests. Note that an F statistic is unitless to yield an area under a curve while RMSE is the same units as our data observations.

Scatter Plot with Line of Best Fit shown

Our AdaBoost model explains 94.6611% of the variation in median value of occupied homes.

Scatterplot of predicted vs. actual gives us a better feel of the improvement between the Linear Regression and AdaBoost Models